to be honest, i dont trade enough spreads in a variety of markets to give any meaningful conclusions. mrci isnt a tipping service though. they provide several trading ideas. it is upto the individual to decide which - if any spreads to enter, as well as when to enter, and when to get out, how to manage the trade etc etc. my general approach is basically tieing up the seasonal window with simple ta entries. you are probably better looking at 'andy's spread trading journal'. andy jordan uses mrci to base his trades on - i believe he uses joe ross techniques (simple 123 & hooks) he has doubled his account in around 6months. http://www.elitetrader.com/vb/showthread.php?s=&threadid=41382 may also want to look at Gold Traders posts - more of a buy & hold approach while adding on at specific points. he uses a stochastic divergence to get in, and uses the parabolic SAR to keep him in.
Fred Thanks I have followed that journal and it is a good one... I was just curious about mrci because I know of a few brokers that claim to be making excellent returns using it and they where encouraging me to give it a try but you know what motivates them ..
I have used MRCI for several years as a platform. The problem is that there is so much 'stuff' to sift through. If you look at the trades, there will be one trade from nov 1 to march 1. Within that longer timeframe trade, you will find many shorter duration trades that are the same. (a prime example that comes to mind is LCJ/LCM or LCQ) In doing analog work on them, I find that the longer trades typically work while the short duration trades are merely the result of coincidental testing. MRCI tests on daily patterns. It cycles through by testing each day against all other possibilities. I expect that is the best way to begin the task of isolating spreads and they tell you in their disclaimer to use their work as a beginning.
I am still trying to get a handle on this. I want to understand so I can get TradersStudio to help people who want to trade spreads. This is my current understanding; let me know constructively if I am on the correct path. 1) Spreads between two contracts on the same market. These trades can only be made on a single contract. This means that we need to be able to produce consolidated results across a basket, in which the basket is create from individual contracts. This means we need to fix up issues created by overlapping dates. 2) Two different market, which we trade as a spread. These can have weighting factors for example 3 contract of market A to 1 contract of market B. These types of spreads would be a candidate for some new type of continuous data, like we have already done for individual stocks in TradersStudio, to solve the problem of split adjustment. I think if we can solve these issues we can begin to develop 100% mechanical systems for spread trading. We can also address seasonal spreads using walk forward methodologies so that we don't have the problems caused by the benefits of hindsight which most seasonal have, for example just because something happens 18 of the past 20 years does not mean that you would have actually taken these trades. This is because the ten year before this period might have had the opposite pattern and you would have not reached 80% bias until the past four years. Hindsight fools you into thinking you would have really made this money. The only way to address seasonality is with walk forward seasonal.
"I think if we can solve these issues we can begin to develop 100% mechanical systems for spread trading." GOOD LUCK, LOL
If you guys can program as well as you say you can write a program that scalps between the "spread price" and price of the legs. Often with Eurodollars you only need to get filled "gaining the bid/ask" on one contract and you can punch on the other two and have a built in profit. I have done this manually with a bit of success but it is very boring and much better suited for automation with a watchful eye. I would not worry about rollover or back testing thats for sure.
Use an autospreader (TT / Ecco) to do this.... I don't think this should be your bread and butter strategy though... more like an extra...